When reduction strategies are put in place and mental health consumers are still secluded: An analysis of clinical and sociodemographic characteristics

Internationally, seclusion practices continue to be the subject of intense clinical health service and academic scrutiny. Despite extensive efforts to reduce and eliminate this controversial practice, seclusion remains a clinical intervention widely used in contemporary mental health service settings. Early identification of people who are at risk for seclusion and the timely application of alternative evidence-based interventions are critical for reducing incidents of seclusion in real-world practice settings. This retrospective study aimed to determine the relationship between sociodemographic and clinical characteristics, and the use of seclusion for those mental health consumers for whom evidence-based seclusion-reduction initiatives had little impact. A 12-month centred moving average was fitted to seclusion data from a psychiatric inpatient unit over 2 years to determine stabilization in seclusion reduction. The number of consumers admitted was calculated from the point of stabilization for 1 year (n=469). In this cohort, univariate analysis sought to compare the characteristics of those who were secluded and those who were not. A multivariate logistic regression model was undertaken to associate future seclusion based on significant independent variables. Of those people admitted, 88 (19%) were secluded. The majority of seclusions occurred in the first 5 days (70/88, 79%). Multivariate logistic regression indicated that three variables maintained their independent associative risk of seclusion: (i) age less than 35 years; (ii) assessment of risk of violence to others; and (iii) a history of seclusion. The implications of these findings for nursing practice are discussed.